Saturday, April 08, 2006

Clean insulators are a man’s best friend

Like all ICP mass spectrometers, our machine spews a lot of ionized gunk into the extraction lens and first lens stack, so after a couple of weeks of operation, these tend to get a little bit grimy. Usually this is no big deal. We swap in the clean sets, clean the dirty ones, and continue on our merry way. But halfway through last month, this technique stopped working.

The clean set of lenses was running at bizarre voltages, and the sensitivity was rather ordinary. So, we cleaned them again, but to no avail. In fact, we ended up purposely running the old, dirty set while we tried to figure out why the clean lenses were not working.

It turns out that taking the gunk off the lenses was necessary, but not sufficient, to restore the machine to its usual happy state. The insulators also had to be replaced, since they had gunked up as well.

So, dirty lens gunk seems to have the bizarre quality that it is insulative enough to build up a static charge sufficient to deflect the beam, but conductive enough to let voltage leak off the lenses. It just takes insulators a lot longer to gunk up to the point where they stop functioning properly, so we can get out of the habit of checking them. This just goes to prove that Murphy’s Law applies to dirt.

I have no idea what this gunk actually is. Most of the material that gets aspirated into the mass spectrometer is argon, which is well known for its reluctance to take solid forms. I suppose that if I really wanted to know what it was, I could just whack a spare dirty lens in the machine and blast away at the gunk deposits. But that would damage the lens more than the gunk does, since craters aren’t easily removed with a bit of polishing compound and some elbow grease.

Besides, it doesn’t really matter what it is, as long as we can prevent it from building up to the point where the machine stops working. It’s not like that knowledge would allow us to scrub less vigorously in order to remove it. So that question, for now, will continue to remain unknown.

No comments: